Pediatric Celiac Disease Patients Show Alterations of Dendritic Cell Shape and Actin Rearrangement

International Journal of Molecular Sciences

Article Pediatric Celiac Disease Patients Show Alterations of Dendritic Cell Shape and Actin Rearrangement

Valentina Discepolo 1,†, Giuliana Lania 1,†, Maria Leonarda Gertrude Ten Eikelder 2, Merlin Nanayakkara 1, Leandra Sepe 3, Rossella Tufano 3, Riccardo Troncone 1, Salvatore Auricchio 1, Renata Auricchio 1, Giovanni Paolella 1 and Maria Vittoria Barone 1,*

1 European Laboratory for the Investigation of Food Induced Diseases, Department of Translational Medical Science, Section of Pediatrics, and ELFID, University Federico II, Via S. Pansini 5, 80131 Naples, Italy; [email protected] (V.D.); [email protected] (G.L.); [email protected] (M.N.); [email protected] (R.T.); [email protected] (S.A.); [email protected] (R.A.); [email protected] (G.P.) 2 Department of Gynecology, Leiden University Medical Centre, Albinusdreef 2, 2333 ZA Leiden, The Netherlands; [email protected] 3 Department of Molecular Medicine and Medical Biotechnologies, University of Naples Federico II, Via S. Pansini 5, 80131 Naples, Italy; [email protected] (L.S.); [email protected] (R.T.) * Correspondence: [email protected] † These two authors contributed equally to this work.

Abstract: Celiac disease (CD) is a frequent intestinal inflammatory disease occurring in genetically susceptible individuals upon gluten ingestion. Recent studies point to a role in CD for genes involved in cell shape, adhesion and actin rearrangements, including a Rho family regulator, Rho GTPase- activating protein 31 (ARHGAP31). In this study, we investigated the morphology and actin cy- Citation: Discepolo, V.; Lania, G.; toskeletons of peripheral monocyte-derived dendritic cells (DCs) from children with CD and controls Ten Eikelder, M.L.G.; Nanayakkara, when in contact with a physiological substrate, fibronectin. DCs were generated from peripheral M.; Sepe, L.; Tufano, R.; Troncone, R.; blood monocytes of pediatric CD patients and controls. After adhesion on fibronectin, DCs showed a Auricchio, S.; Auricchio, R.; Paolella, higher number of protrusions and a more elongated shape in CD patients compared with controls, G.; et al. Pediatric Celiac Disease as assessed by immunofluorescence actin staining, transmitted light staining and video time-lapse mi- Patients Show Alterations of Dendritic Cell Shape and Actin croscopy. These alterations did not depend on active intestinal inflammation associated with gluten Rearrangement. Int. J. Mol. Sci. 2021, consumption and were specific to CD, since they were not found in subjects affected by other in- 22, 2708. https://doi.org/10.3390/ testinal inflammatory conditions. The elongated morphology was not a result of differences in ijms22052708 DC activation or maturation status, and did not depend on the human leukocyte antigen (HLA)- DQ2 haplotype. Notably, we found that ARH-GAP31 mRNA levels were decreased while RhoA-GTP Academic Editor: Roberto De Giorgio activity was increased in CD DCs, pointing to an impairment of the Rho pathway in CD cells. Accord- ingly, Rho inhibition was able to prevent the cytoskeleton rearrangements leading to the elongated Received: 3 February 2021 morphology of celiac DCs upon adhesion on fibronectin, confirming the role of this pathway in the Accepted: 24 February 2021 observed phenotype. In conclusion, adhesion on fibronectin discriminated CD from the controls’ DCs, Published: 8 March 2021 revealing a gluten-independent CD-specific cellular phenotype related to DC shape and regulated by RhoA activity. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in Keywords: celiac disease; dendritic cells; actin cytoskeleton; cell shape; adhesion; fibronectin; published maps and institutional affil- RhoA; ARHGAP31 iations.

1. Introduction Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. Celiac disease (CD) is characterized by activation of both the adaptive and the innate This article is an open access article immune responses upon gluten ingestion. Some gliadin peptides (i.e., A-gliadin P57- distributed under the terms and 68) are presented by antigen-presenting cells, in the context of human leukocyte antigen conditions of the Creative Commons (HLA)-DQ2 and/or -DQ8 molecules after deamidation by the enzyme tissue transglu- Attribution (CC BY) license (https:// taminase (tTg). The presentation of these peptides by antigen-presenting cells, such as creativecommons.org/licenses/by/ dendritic cells (DCs), is able to induce a gluten-speciﬁc adaptive T helper (Th)-1 inﬂam- 4.0/). matory response [1]. Other gliadin peptides (i.e., P31–43) are able to initiate an innate

Int. J. Mol. Sci. 2021, 22, 2708. https://doi.org/10.3390/ijms22052708 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2021, 22, 2708 2 of 16

immune response [2,3]. The intestinal mucosal damage typical of CD is characterized by infiltration of intraepithelial lymphocytes, proliferation of crypt enterocytes as an early alteration of CD mucosa resulting in crypt hyperplasia [4], and ultimately villous atrophy mediated by cytotoxic T cells and inflammatory cytokines. These alterations are reverted upon gluten withdrawal. Modifications of the actin cytoskeleton [5] and gluten-dependent shortening of enterocytes have been found in CD patients [6]. Alterations of the tight junctions (TJ) regulating intestinal permeability persisted even in patients on a gluten-free diet (GFD) with a normalized intestinal architecture [7–11]. Moreover, a disruption of the intestinal epithelium integrity was found in early-stage CD [12]. Polymorphisms in the TJ genes Par-3 family cell polarity regulator (PARD3) and membrane-associated guanylate kinase, WW and PDZ domain-containing 2 (MAGI2) have been associated with CD susceptibility [13]. Interest- ingly, the protein phosphatase 2 regulatory subunit B alpha (PPP2R3A) gene, implicated in the negative control of cell growth, division and TJ regulation, remains down-regulated at the intestinal level in patients on a GFD [14]. These observations suggest a role for cellular structural alterations in the pathogenesis of CD. Of note, gliadin peptides, including P31–43, can interfere with actin rearrangements in both the CD mucosa and epithelial cell lines [9,15–20]. More recent studies predicted a role in CD pathogenesis for lipoma-preferred partner (LPP), C1ORF106 (C1 Orfan 106), Rho GTPase-activating protein 31 (ARHGAP31) and protein tyrosine phosphatase receptor type K (PTPRK) genes, which play a role in actin cytoskeleton rearrangement, cell–cell adhesion and in epidermal growth factor (EGF)/ EGF receptor (EGFR) pathway activation [21–24]. Transcripts of ARHGAP31, a Rho-GTPase- activating protein (GAP) which is required for cell spreading, polarized lamellipodia formation and cell migration, are altered in CD immune cells [24]. The Rho GTPases control the organization of the cytoskeleton, cell migration, cell–cell and cell–matrix adhesion, cell cycle progression, gene expression and cell polarity. Interestingly, RhoB was found to be involved in the intestinal remodeling of CD intestinal mucosa [25] and genetic variations of miosin IX B (MYO9B), which serves as a regulator of the Rho-dependent signaling pathways, have been associated with CD [26]. The link between these genes and a cellular phenotype has been studied only for LPP. In fact, alterations of LPP sub-cellular distribution, together with modifications of cell shape, the actin cytoskeleton and focal adhesions, have been described in fibroblasts from CD patients [27]. These data confirmed that structural cellular alterations, mainly involving the actin cytoskeleton, are present in cells from CD patients. Whether similar structural alterations can be observed also in circulating immune cells from CD patients remains to be defined. In this study, we aimed to address this issue by investigating cell shape and the related cytoskeleton alterations in monocyte-derived DCs from children with CD compared with controls. The choice fell on peripheral blood monocyte-derived DCs since they are disjoined by the intestinal inflammatory milieu and are easily accessible, yet representing a cell subset of key immunological relevance in CD. To verify the hypothesis that structural cytoskeleton alterations are present in peripheral immune cells from CD patients, we studied the morphological characteristic of DC shape upon adhesion on fibronectin using three distinct assays and investigated their link with RhoA activity. We showed that DCs from CD patients presented lower levels of ARHGAP31, increased Rho A activation and displayed an elongated morphology when in contact with fibronectin compared with controls. Thus, adhesion on fibronectin revealed a “CD cellular phenotype” of DCs independent of gluten or HLA-DQ2 and regulated by Rho A activity.

2. Results 2.1. Rho-GTP and ARHGAP Levels in CD DCs DCs were generated from peripheral blood monocytes [28] of pediatric celiac patients on a gluten-containing diet (GCD-CD) in the active phase of the disease and on GFD (GFD-CD) in the remission phase of the disease, as well as those of healthy controls. Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 3 of 17

Int. J. Mol. Sci. 2021, 22, 2708 2.1. Rho-GTP and ARHGAP Levels in CD DCs. 3 of 16 DCs were generated from peripheral blood monocytes [28] of pediatric celiac patients on a gluten-containing diet (GCD-CD) in the active phase of the disease and on GFD (GFDSmall- GTP-bindingCD) in the remission proteins phase of the of Rho the familydisease, are as keywell regulators as those of of healthy cytoskeletal controls. dynamic Small GTPand- affectbinding many proteins cellular of processes,the Rho family including are key cell regulators polarity, migration,of cytoskeletal vesicle dynamic trafficking and affectand cytokine many cellular production. processes The, Rhoincluding family cell of polarity, GTPases migration, in particular vesicle has beentrafficking shown and to cytokineregulate cellproduction. shape and The lamellipodia Rho family of production GTPases in in par bloodticular cells has [ 29been,30 ].shown This to prompted regulate cellus to shape test RhoAand lamellipodia activity in DCs production from GCD-CD in blood patients cells [29,30 and]. controls This prompted when plated us to on test fi- RhoAbronectin, activity a physiological in DCs from substrate.GCD-CD patients In Figure and1A,B controls and Supplementary when plated on Figure fibronectin, S1A,B, a Westernphysiological blot analysissubstrate. of In RhoA-GTP, Figure 1A, theB and activated Supplementary form of RhoA, Figure is S shown.1A,B, W RhoA-GTPestern blot analysislevels were of RhoA higher-GTP, in GCD-CD the activated DCs with form respect of RhoA, to controls. is shown. In RhoA fact, the-GTP ratio levels between were higherthe active in GCD form-CD of Rho DCs and with the respect total protein to controls. (RhoA-GTP/Rho) In fact, the ratio was between significantly the active lower form in ofcontrols Rho and (0.03 the± total0.04) protein compared (RhoA with-GTP/Rho) GCD-CD patientswas significantly (0.56 ± 0.17, lowerp = 0.0001),in controls indicating (0.03 ± 0.0that4) incompared active CD with patients, GCD- anCD activation patients ( of0.56 RhoA ± 0.1 occurred7, p = 0.0001 in DCs), indicating upon contact that in with active the CDsubstrate patients fibronectin, an activation (Figure of1 BRhoA). Nocodazole, occurred in an DCs inductor upon of contact RhoA activity,with the was substrate used as fi- a positivebronectin control (Figure to 1 enhanceB). Nocodazole RhoA-GTP, an inductor levels in controlof RhoA DCs activity, (0.69 ±was0.022 usedvs. as0.03 a positive± 0.04, controlp = 0.0004 to ).enhance Interestingly, RhoA-GTP a difference levels in incontrol RhoA-GTP DCs (0.69 levels ± 0.0 was22 vs also. 0.0 found3 ± 0.04, in p DCs = 0.0004 from). Interestingly,GFD-CD (0.29 a ±difference0.1 vs. 0.03 in RhoA± 0.04-GTPp = 0.0007levels) was compared also found with in controls DCs from (Figure GFD1-C,DCD (and0.29 ±Supplementary 0.1 vs. 0.03 ± 0.0 Figure4 p = S1C,D0.0007),) compared suggesting with that controls RhoA activation (Figure 1C was,D and independent Supplementary of dis- Figureease activity. S1C,D), suggesting that RhoA activation was independent of disease activity. Figure 1

CTR

A UN NOC GCD-CD B

Rho GTP o

Total Rho T

UN NOC GCD-CD

CTR

C CTR GFD-CD D

o Rho GTP h

Total Rho /

CTR GFD-CD E

Figure 1. RhoRho-GTP-GTP and and Rho GTPase GTPase-activating-activating protein ( (ARHGAP)ARHGAP) levels in dendritic cells (DCs)(DCs) from celiacceliac diseasedisease (CD)(CD) patients patients and and controls. controls. Ras Ras homolog homolog family family member member A (RhoA)-GTPA (RhoA)-GTP levels levels are arehigher higher in DCs in DCs from from CD patientsCD patients than than controls controls (CTR) (CTR) after after seeding seeding on fibronectin on fibronectin for 1 h.for 10 16 h.DCs 106were DCs platedwere plated for each for subjecteach subject and each and condition. each condition (A). Representative. (A). Representative Western W blotestern (WB) blot of (WB) RhoA of activity RhoA (RhoA)-GTPactivity (RhoA) levels-GTP in levels dendritic in dendritic cells (DCs) cells from (DCs) one from control one (CTR) control and (CTR) one celiacand one patient celiac on patient a gluten- on a gluten-containing diet (GCD-CD). DCs from CTR were analyzed before (untreated, UN) and after containing diet (GCD-CD). DCs from CTR were analyzed before (untreated, UN) and after treatment treatment with nocodazole (NOC). (B). Densitometric analysis of WB performed as in A from five with nocodazole (NOC). (B). Densitometric analysis of WB performed as in A from five CTRs and five CTRs and five GCD-CD. NOC was used as a positive control on DCs from three CTRs. Compared withGCD-CD. UN DCs NOC from was CTR useds, NOC as apositive promoted control RhoA on activation DCs from, reaching three CTRs. similar Compared levels to those with observed UN DCs infrom GCD CTRs,-CD. NOC Horizontal promoted lines RhoA represent activation, the mean reaching; bars similar, the standard levels to deviation. those observed Ordinary in GCD-CD. one-way Horizontal lines represent the mean; bars, the standard deviation. Ordinary one-way ANOVA with Tukey’s multiple comparison correction was performed; *** p < 0.001. (C). Representative WB showing

Int. J. Mol. Sci. 2021, 22, 2708 4 of 16

RhoA-GTP levels in DCs from a CTR and a gluten-free diet (GFD-CD) patient. (D). Densitometric analysis of WB performed as in C from ﬁve subjects for each group. Horizontal lines represent the mean; bars, the standard deviation. Student’s t-test; * p < 0.05, ** p < 0.01. (E). Quantitative PCR analysis of Rho GTPase-activating protein 31 (ARHGAP31) mRNA levels showing decreased ARHGAP31 in GCD-CD DCs with respect to CTR. The number of subjects analyzed is indicated. The horizontal lines represent the median; bars, the standard deviation. Student’s t-test; ** p < 0.01.

We also evaluated transcriptional levels of ARHGAP31, a Rho-GTPase activating protein, in DCs from GCD-CD patients (Figure1E), showing a significant decrease in ARHGAP31 mRNA in GCD-CD compared with controls’ DCs (p < 0.01). The reduction of a negative regulator of Rho is in line with the increase in Rho activity; together, these data support the induction of the RhoA pathway in peripheral monocyte-derived DCs of CD patients.

2.2. DC Shape in CD Patients Shows a Different Shape Compared with Controls When Adhering on Fibronectin The increase in Rho activity in CD patients’ DCs prompted us to investigate the shape of DCs from GCD-CD and GFD-CD patients in comparison with controls. To highlight the morphology and actin cytoskeleton of DCs upon interaction with fibronectin, fluorescein isothiocyanate (FITC)-conjugated phalloidin immunofluorescent staining was performed (Figure2A). DCs from controls, GCD-CD and GFD-CD patients were seeded on coverslips pre-coated with fibronectin and incubated. Phalloidin staining revealed differences in the shape of DCs from active CD patients with more numerous and long dendrites compared with controls. To quantify the differences in morphology, we measured the length of the phalloidin-stained DCs using the Zeiss LSM 510 software. The maximum stretch of at least 20 DCs per patients was measured, and the average extension was calculated and plotted (Figure1B). DCs from GCD-CD patients ( 79.12 ± 19.96 µm) were significantly longer than control DCs (44.20 ± 17.21 µm; p = 0.0004). GFD-CD DCs also showed an elongated shape (54.88 ± 20.17 µm), although not statistically significant with respect to the controls’ DCs. To further evaluate the altered morphology of DCs, we stained them with crystal violet and observed them with a transmitted light microscope after l hour (1 h) and 3 h adhesion on fibronectin (Figure2C). DCs from CD patients showed a different shape upon fibronectin interaction when compared with controls. DCs showing more than three small dendrites and/or one long dendrite were considered to have an elongated shape. Dendrites were defined as “small” when they were shorter than half of the cellular body length, and “long” when they were longer than half of the cellular body length. Adhering cells were pictured and counted. The percentage of elongated DCs among adhering cells was evaluated and CD patients were found to have a statistically significant higher percentage of elongated DCs compared with controls (Figure2D). After 3 h seeding on fibronectin, 71% ± 6.7 DCs from GCD-CD patients and 57% ± 6.6 DCs from GFD-CD patients showed an elongated shape compared with 36.95% ± 12.4 in the control group (p < 0.001, p < 0.01, respectively,). After l h seeding on fibronectin, only DCs from active CD patients showed a significantly more elongated shape compared with controls (p < 0.001). Int. J. Mol. Sci. 2021, 22, 2708 5 of 16 Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 5 of 17

A CTR GFD CD GCD CD

B 150 *** Controls * GCD CD

m GFD CD ∝ 100

50 cell length

s D D ol C C tr D n CD F Co G G

C CTR GFD CD GCD CD

D 100 **** ** 80 ***

20 percent of elongated among all adherent among all cells 0

(>3 and/or small 1 long dendrite) R D D R D D T C C T C C C D C D F F G G

1h 3h Figure 2. Elongated shape of DCs from CD patients and controls after spreading on fibronectin. (A) Figure 2. Elongated shape of DCs from CD patients and controls after spreading on fibronectin. Representative images of fluorescein isothiocyanate (FITC)-conjugated phalloidin staining high- (lightingA) Representative the actin cytoskeleton images of of DCs fluorescein from CTR, isothiocyanate GFD and GCD CD (FITC)-conjugated subjects. DCs were phalloidin seeded on staining highlightingcoverslips pre-coated the actin with cytoskeleton fibronectin at of a DCs concentration from CTR, of GFD2.5 × 10 and5 cells/mL. GCD CD Experiments subjects. were DCs run were seeded onin duplicate. coverslips Five pre-coated random pictures with fibronectin from each duplicate at a concentration were evaluated. of 2.5 Scale× 10bar5 =cells/mL. 7 µm. (B) ExperimentsThe were run in duplicate. Five random pictures from each duplicate were evaluated. Scale bar = 7 µm. (B) The maximum stretch of at least 20 DC per patients was measured and the average extension

has been calculated and plotted. Quantitative analysis of DCs length, from 17 controls, 8 GCD-CD Int. J. Mol. Sci. 2021, 22, 2708 6 of 16

patients and 10 GFD-CD patients, after seeding on fibronectin for 3 h. Horizontal lines represent the mean; bars, the range. Ordinary one-way ANOVA with Tukey’s multiple comparison correction; * p < 0.05, *** p < 0.001. (C) Light microscopy images of fibronectin-adhering DCs from CTR, GFD-CD and GCD-CD patients spread on fibronectin for l (upper panel) and 3 h (lower panel). Crystal violet staining was performed. Scale bar = 10 µm. Each experiment was performed in duplicate, with 105 DCs plated in each well. At least two independent experiments for each subject were performed. Six pictures were taken for each duplicate. Representative images are shown. (D) The percentage of fibronectin-adhering DCs showing an elongated shape, defined as a cell with more than three short and/or one long dendrites, among all spreading cells was assessed in five CTR, four GFD-CD and four GCD-CD. At least 50 cells per subject were counted. Scale bar = 10 µm. The horizontal lines represent the mean; bars, the standard deviation. Ordinary one-way ANOVA with Sidak’s multiple comparison test; ** p < 0.01, *** p < 0.001, **** p < 0.0001.

2.3. DC Shape in Patients with Food Allergies or Inflammatory Bowel Diseases and Healthy Subjects Evaluated by Immunofluorescence Staining To assess whether the elongated morphology described in DCs from CD patients was also present in other diseases and to explore whether intestinal inflammation could contribute to altering DC shape, we evaluated the morphology and actin cytoskeleton of DCs from patients with inflammatory bowel diseases and food allergies. DCs from patients with Crohn’s disease, ulcerative colitis and food allergies, were seeded on coverslips pre- coated with fibronectin and analyzed by confocal microscopy upon a FITC-conjugated phalloidin immunofluorescent staining to highlight the actin cytoskeleton (Figure3A). None of the analyzed patients affected by diseases other than CD showed the specific elongated morphology observed in CD DCs, suggesting that the observed morphological changes are specific to CD. Cytoskeleton alterations characterized by increased membrane ruffling (Figure3A) have been described in DCs from Crohn’s disease patients and the related to inflammation [31]. Importantly, the inhibition of RhoA activation by treatment with the Rho-associated protein kinase (ROCK) inhibitor Y-27632 was able to prevent DC shape alterations in CD (Figure3B), demonstrating that the elongated shape of DCs from CD patients was dependent on RhoA activity. Moreover, to exclude the possibility that HLA-DQ2-predisposing alleles could impact the observed morphology of DCs in CD patients, we investigated DC shape in control subjects carrying or not carrying the HLA-DQ2 haplotype and did not observe any difference (Supplementary Figure S3).

2.4. DC Shape in CD Patients and Healthy Subjects Evaluated by Video Time-Lapse Microscopy Monocytes derived DCs from active CD patients and healthy subjects were plated on fibronectin-coated wells and filmed live for 3 h by video time-lapse microscopy to observe their ability to interact with the substrate (see Supplemental Movies S1 and S2). DCs were recorded while spreading on fibronectin. Before spreading, DCs appeared as small translucent circles, while upon spreading on the substrate, they appeared less translucent and started to project dendrites. Within 2 h, dendrites were formed and could retreat. DCs from CD patients showed longer dendrites with respect to the controls at any time observed. Images were taken every 18 min. In Figure4, images taken every 36 min from two representative subjects are shown. The ability of the DCs from CD subjects to adhere on the substrate was not impaired; in fact, the number of adhering cells after seeding on fibronectin was the same as in the control (Supplementary Figure S2). Int.Int. J.J. Mol.Mol. Sci.Sci. 20212021,, 2222,, 2708x FOR PEER REVIEW 77 ofof 1617

Control Celiac Disease Food Allergy Ulcerative Colitis Crohn’s disease A

Untreated Y-27632 Untreated Y-27632 B

CTR GCD-CD FigureFigure 3.3. DCDC shapeshape andand cytoskeletoncytoskeleton evaluatedevaluated byby immunofluorescenceimmunofluorescence staining.staining. ((AA)) RepresentativeRepresentative immunofluorescenceimmunofluorescence imagesimages ofof FITC-conjugatedFITC-conjugated phalloidinphalloidin staining,staining, highlightinghighlighting thethe actinactin cytoskeletoncytoskeleton ofof fibronectin-adheringfibronectin-adhering DCsDCs fromfrom oneone controlcontrol andand patientspatients withwith GCD-CD,GCD-CD, foodfood allergyallergy oror inflammatory inflammatory bowelbowel diseasedisease (ulcerative(ulcerative colitiscolitis andand Crohn’sCrohn’s disease),disease), showing that the elongated morphology of DCs was specific to CD. FITC-conjugated phalloidin staining was performed showing that the elongated morphology of DCs was specific to CD. FITC-conjugated phalloidin staining was performed on on DCs plated on coverslips pre-coated with fibronectin at a concentration of 2.5 × 105 cells/mL. One representative image DCs plated on coverslips pre-coated with fibronectin at a concentration of 2.5 × 105 cells/mL. One representative image out of at least three subjects per group is shown. Each experiment was done in duplicate; six pictures were taken for each outduplicate. of at least Scale three bar = subjects 7 µm. (B per) Representative group is shown. immunofluorescence Each experiment images was done of phalloidin in duplicate; staining six pictures of DCs from were four taken GCD- for eachCD patients duplicate. and Scale controls bar = 7platedµm. (Bon) Representativecoverslips pre-co immunofluorescenceated with fibronectin images at a ofconcentration phalloidin staining of 2.5 × of 10 DCs5 cells/mL from four and GCD-CDincubated patients at 37 °C and for controls3 h before plated (Untreated) on coverslips and after pre-coated treatment with with fibronectin the Y-27632 at Rho-associ a concentrationated protein of 2.5 × kinase105 cells/mL (ROCK) andinhibitor. incubated Treatment at 37 ◦withC for Y-27632 3 h before prevented (Untreated) protrusion and after form treatmentation and withcell shape the Y-27632 alterations Rho-associated observed in protein GCD-CD kinase DCs Int. J. Mol. Sci. 2021, 22, x FOR PEER REVIEW 8 of 17 (ROCK)(right panels), inhibitor. confirming Treatment that with the Y-27632 DC shape prevented of CD patients protrusion depended formation on andRhoA cell activation. shape alterations observed in GCD-CD DCs (right panels), confirming that the DC shape of CD patients depended on RhoA activation. 2.4. DC Shape in CD Patients and Healthy Subjects Evaluated by Video Time-Lapse Microscopy Monocytes derived DCs from active CD patients and healthy subjects were plated on fibronectin-coated wells and filmed live for 3 h by video time-lapse microscopy to observe their ability to interact with the substrate (see Supplemental Movies 1 and 2). DCs were recorded while spreading on fibronectin. Before spreading, DCs appeared as small translucent circles, while upon spreading on the substrate, they appeared less translucent and started to project dendrites. Within 2 h, dendrites were formed and could retreat. DCs from CD patients showed longer dendrites with respect to the controls at any time observed. Images were taken every 18 min. In Figure 4, images taken every 36 min from two representative subjects are shown. The ability of the DCs from CD subjects to adhere on the substrate was not impaired; in fact, the number of adhering cells after seeding on fibronectin was the same as in the control (Supplementary Figure S2).

Figure 4. Spreading ability of fibronectin-adhering DCs. DCs from three active celiac patients (GCD-CD) and from three Figure 4. Spreading ability of fibronectin-adhering DCs. DCs from three active celiac patients (GCD-CD) and controls were seeded on wells pre-coated with fibronectin at a density of 2.5 × 105 cells/mL, and their spreading properties from three controls were seeded on wells pre-coated with fibronectin at a density of 2.5 × 105 cells/mL, and were recorded live for 3 h to observe dendrite formation. Fifty cells in each frame were followed over time for a total their spreading properties were recorded live for 3 h to observe dendrite formation. Fifty cells in each frame of 200 DCs per patient. One picture every 36 min of the movie is shown, starting from T0 (the time when DCs were were followed over time for a total of 200 DCs per patient. One picture every 36 min of the movie is shown, seeded in the well) until T180 (the time when the film was stopped). Images from a representative CD patient and a starting from T0 (the time when DCs were seeded in the well) until T180 (the time when the film was stopped). representative control out of at least three subjects per group are shown. White arrows indicate the same cells in each panel. Images from a representative CD patient and a representative control out of at least three subjects per group Scale bar = 20 µm. are shown. White arrows indicate the same cells in each panel. Scale bar= 20 µm.

2.5. Evaluation of DC Markers of Maturation in Controls and CD Patients before and after Adhesion on Fibronectin To evaluate whether the different DC shape observed in CD subjects was dependent on immunological activation of peripheral cells, the immunological surface profile of DCs before and after interaction with fibronectin was assessed. Surface expression of cluster of differentiation (CD)14, CD40, CD83, CD86 and CD11c was evaluated by flow cytometry to establish the DC phenotype before interaction with fibronectin (Figure 5A,B). CD14 staining was performed to exclude monocyte contamination; CD86, CD83 and CD40 expression was assessed to determine DCs’ immunological activation and maturation (CD83), while CD11c staining ensured that DCs belonged to the myeloid subset. Notably, DCs from both healthy subjects and CD patients had similar surface marker profiles (Fig- ure 5A,B), consistent with immature (CD83–) non-activated (CD40–, CD86low) and myeloid DCs (CD14–CD11c+). Overnight lipopolysaccharide (LPS) induced maturation of DCs both in controls and in CD subjects (Figure 5A–C), as shown by a statistically significant increase in both CD83 mean fluorescence intensity (Figure 5B) and the percentage of positive cells (Figure 5C) upon LPS treatment compared with untreated cells. (LPS). The ability of fibronectin interaction to induce an alteration of DCs’ surface marker profile was also tested. Double immunofluorescence staining was performed on DCs from active CD patients and healthy controls after adhesion on fibronectin, showing that adhesion on fibronectin did not alter the surface marker profile (HLA-DR+, CD14+ CD83+, CD86+, CD11c+) of DCs from CD patients (Figure 5D). Altogether, these data show that, DCs from both controls and CD patients displayed an immature/adhering phenotype in the presence of fibronectin; however, DCs from CD patients had a more elongated shape.

Int. J. Mol. Sci. 2021, 22, 2708 8 of 16

FigureFigure 5. 5.DCs’ DCs’ surface surface marker marker profile profile of CD of patients CD patients and controls and co beforentrols and before after adhesionand after on adhesion fibronectin. on (fibronectin.A). Flow cytometric(A). Flow analysis cytometric of DCs’ analysis surface of markers DCs’ surface profile beforemarkers interaction profile before with fibronectin. interaction DCs with generated fibronectin. from DCs peripheral gener- monocytesated from wereperipheral used at monocytes a concentration were of used 2.7–8.3 at a× concentration105 cells/mL inof 24-well2.7–8.3 plates× 105 cells/mL and stained in with24-well the plates antibodies and indicatedstained with in the the figure. antibodies Fluorescence indicated intensity in ofthe the figure. analyzed Fluorescence samples is representedintensity of in the a histogram analyzed plot samples with a blackis repre- line (NT,sented not in treated); a histogram green lines plot refer with to a lipopolysaccharide black line (NT, no (LPS)-treatedt treated); green samples; lines pink refer lines to referlipopolysaccharide to the isotype-matched (LPS)- controltreated (ISO). samples; Representative pink lines histogram refer to plots the fromisotype-matche one GCD-CDd patientcontrol and (ISO). one CTRRepresentative are shown. ( Bhistogram,C). Quantification plots from and one GCD-CD patient and one CTR are shown. (B,C). Quantification and statistical analysis of cluster differentiation (CD)83 surface expression as assessed by flow cytometry in DCs from CD patients and CTR showing similar CD83 surface expression levels. For CD83, mean fluorescence intensity (MFI) (B) and the percentage of positive cells (C) are shown. Overnight LPS treatment, used as a positive control, induced a significant increase in CD83 MFI (B) and the percentage of CD83+ cells (C) in both CTR and in CD subjects compared with non-treated (NT) cells. Each dot represents one subject. The horizontal lines represent the median; bars, the standard deviation. Student’s t-test; * p < 0.05, ** p < 0.01, *** p < 0.001. (D). DCs from active CD patients were plated onto coverslips pre-coated with fibronectin at a concentration of 2.5 × 105 cells/mL and incubated at 37 °C. To assess the surface marker profile after 3 h of adhesion on fibronectin by confocal microscopy, DCs were stained with FITC-conjugated anti-human leukocyte antigen (HLA)-DR antibody (green staining in each image) and with a phycoerythrin (PE)-conjugated antibody (red staining) against different surface markers, as indicated in each image. Yellow color represents merged staining. Cells appear round because of the four- degree incubation used to avoid antibody endocytosis. Representative images of at least three subjects done in duplicate are shown. More than 20 cells per patients were observed. Scale bar = 8 µm.

Int. J. Mol. Sci. 2021, 22, 2708 9 of 16

statistical analysis of cluster differentiation (CD)83 surface expression as assessed by flow cytometry in DCs from CD patients and CTR showing similar CD83 surface expression levels. For CD83, mean fluorescence intensity (MFI) (B) and the percentage of positive cells (C) are shown. Overnight LPS treatment, used as a positive control, induced a significant increase in CD83 MFI (B) and the percentage of CD83+ cells (C) in both CTR and in CD subjects compared with non-treated (NT) cells. Each dot represents one subject. The horizontal lines represent the median; bars, the standard deviation. Student’s t-test; * p < 0.05, ** p < 0.01, *** p < 0.001. (D). DCs from active CD patients were plated onto coverslips pre-coated with fibronectin at a concentration of 2.5 × 105 cells/mL and incubated at 37 ◦C. To assess the surface marker profile after 3 h of adhesion on fibronectin by confocal microscopy, DCs were stained with FITC-conjugated anti-human leukocyte antigen (HLA)-DR antibody (green staining in each image) and with a phycoerythrin (PE)-conjugated antibody (red staining) against different surface markers, as indicated in each image. Yellow color represents merged staining. Cells appear round because of the four-degree incubation used to avoid antibody endocytosis. Representative images of at least three subjects done in duplicate are shown. More than 20 cells per patients were observed. Scale bar = 8 µm.

Overnight lipopolysaccharide (LPS) induced maturation of DCs both in controls and in CD subjects (Figure5A–C), as shown by a statistically significant increase in both CD83 mean fluorescence intensity (Figure5B) and the percentage of positive cells (Figure5C) upon LPS treatment compared with untreated cells. (LPS). The ability of fibronectin interaction to induce an alteration of DCs’ surface marker profile was also tested. Double immunofluorescence staining was performed on DCs from active CD patients and healthy controls after adhesion on fibronectin, showing that adhesion on fibronectin did not alter the surface marker profile (HLA-DR+, CD14+ CD83+, CD86+, CD11c+) of DCs from CD patients (Figure5D). Altogether, these data show that, DCs from both controls and CD patients displayed an immature/adhering phenotype in the presence of fibronectin; however, DCs from CD patients had a more elongated shape.

3. Discussion In this study, we analyzed the morphology and actin rearrangements of monocyte- derived DCs of pediatric CD patients, using three different techniques, namely immunofluorescence, transmitted light microscopy upon crystal violet staining and video time-lapse microscopy, showing the presence of structural alterations of CD cells, defining a CD-specific cellular phenotype. Importantly, we linked this phenotype to increased RhoA activity and to decreased ARHGAP transcriptional levels, a Rho-GAP that genetic studies have found to be associated with CD pathogenesis. Alterations of the actin cytoskeleton and tight junctions have previously been described in CD enterocytes, indicating that cytoskeletal rearrangements and cell contacts might be involved in the CD lesions [5–12]. An increase in intestinal permeability resulting from a loss of the barrier function due to tissue damage is a constant feature in CD subjects [32]. GFD can partially or fully restore these alterations [11,33]. In this paper, we show an alteration of the morphology and actin cytoskeleton of DCs in pediatric CD patients. Adhesion on fibronectin induced the formation of elongated dendrites in a higher number of DCs from CD patients with respect to controls, indicating that CD DCs reacted differently upon interaction with a physiological substrate. This altered shape upon contact with the substrate was displayed by DCs from both treated and untreated pediatric CD patients, revealing a “CD cellular phenotype” independent of dietary gluten content. This phenotype is specific to children with CD, as it was absent in DCs from pediatric patients with Crohn’s disease, ulcerative colitis and food allergies. Moreover, this CD-specific phenotype does not depend on the disease activity, since it could also be observed in children on a GFD and did not seem to depend on the presence of predisposing HLA-DQ2 alleles, since it was not observed in controls carrying this haplotype. Furthermore, DCs from both CD pediatric patients and controls had the same surface immunological profile before and after adhesion on fibronectin, correspond- ing to an immature, non-activated and adhering phenotype of myeloid DCs. A similar CD cellular phenotype characterized by altered cell shape and actin distribution has also been found in skin fibroblasts from CD patients, another cell population located far from Int. J. Mol. Sci. 2021, 22, 2708 10 of 16

the intestinal lesion and in the absence of gliadin peptides [27], suggesting a structural constitutive alteration. In this paper, we also investigated the differences in the activity of RhoA, a protein able to regulate cell shape and dendrite formation in monocytes [34]. Interestingly, we found that in comparison with controls, upon contact with fibronectin, RhoA-GTP levels were increased in DCs from pediatric CD patients and that the inhibition of RhoA activity prevented the morphological alterations resulting in an elongated shape observed in DCs from CD patients. Altogether, these data point to an alteration of the RhoA pathway in CD DCs. Accordingly, together with an increased RhoA activity, we observed a decrease in ARHGAP31 transcript levels. The reduction of a negative regulator of Rho is, in fact, com- patible with increased Rho activity. In line with our findings, genetic studies on peripheral blood monocytes point to an altered expression of ARHGAP31 in CD. This observation could suggest that the CD cellular phenotype described in this study could be associated with a “cell type-specific expression” of CD-related genes, including ARHGAP31 [24]. Constitutive alterations of CD cells, in the absence of a gliadin challenge, have been described in different systems and both in vivo and in vitro assays [35–38]. These alterations point to a derangement of signaling, proliferation and the cytoskeleton, leading to inflammation, innate immunity activation and cellular stress in the enterocytes and fibroblasts of CD patients. An impairment of the endocytic trafficking could contribute to explain- ing these structural alterations [36]. More recently, ex vivo experiments in intestinal organoids [39,40] demonstrated that staminality and inflammation pathways are altered in GFD-CD organoids even before gluten exposure. Moreover, gluten challenge further impacts the same pathways already altered in GFD biopsies [37] and organoids [39]. Taken together, these observations describe a “CD cellular phenotype” characterized by altered cell shape, actin remodeling, signaling and innate immunity activation, suggesting that even in polygenic disorders, a “disease-specific cellular phenotype” can be observed in almost all pediatric subjects. Further studies will clarify whether these same alterations can be observed in adults with CD. Similarly, a disease-specific cellular phenotype, independent of disease-associated gene variants, has been observed in another polygenic intestinal inflammatory diseases, such as Crohn’s disease. Indeed, Paneth cells from Crohn’s intestinal biopsies display an impairment in autophagy regardless of the presence of genetic variants of the autophagy pathway, which have been detected only in few Crohn’s patients [41]. In conclusion, we showed that adhesion on fibronectin revealed a different morphology of DCs from pediatric CD patients compared with controls. This altered DC shape may be considered as part of the gluten-independent alterations of CD cells defining a “CD cellular phenotype”, characterized not only by alterations of the signaling and inflammatory markers, but also by structural cellular alterations [21,36,42–44].

4. Materials and Methods 4.1. Patients Monocyte-derived DCs were obtained from pediatric CD patients on a GCD in the active phase of the disease (37 patients, mean age 6.6 years), or on a GFD in the remission phase of the disease (11 subjects, mean age of 11 years) and from controls, including subjects affected by gastroesophageal reflux (31 subjects, mean age of 7.5 years). GCD-CD patients included children with positive serology, including anti-tTg immunoglobulin A (IgA) antibodies (>50 U/mL) and anti-endomysium antibodies (EMA), and displaying duodenal villous atrophy (Marsh T3) at the small intestinal biopsy (Table1). Patients with GFD-CD had negative CD-specific serology, including anti-tTg IgA antibody titers ranging between 0 and 1.5 U/mL, negative EMA and a normal duodenal biopsy (Marsh score T0). Patients had been on a GFD for at least 4 years prior to enrollment; the adherence to the GFD was complete and patients had remission of the symptoms and negative CD-specific serology. Anti-tTg antibody titers were measured using the Eurospital kit EU-tTG (Trieste, Italy). To compare our observations with patients with other intestinal inflammatory conditions, Int. J. Mol. Sci. 2021, 22, 2708 11 of 16

we also enrolled four pediatric patients with food allergies associated with intestinal symptoms and 6 with pediatric inﬂammatory bowel diseases, particularly 3 subjects with Crohn’s disease and 3 with ulcerative colitis. All enrolled subjects were children and their parents provided written informed consent for the use of their blood cells in this study. The protocol for this study was approved by the Ethical Committee of the University “Federico II”, Naples, Italy (Ethical Committee approval, Protocol no. 230/05).

Table 1. Clinical data of patients and controls.

Anti- Mean Age Biopsy Marsh Serum Anti-TG2 Patients Sex Endomysium (Years) Classiﬁcation * Antibodies (U/mL) Antibodies (EMA) Controls (N = 31) 7.5 M = 17 F = 14 T0 0–1.5 Negative T3a = 0 GCD-CD (N = 37) 6.6 M = 12 F = 25 >50 Positive T3b/c = 16 T3c = 21 GFD-CD (N = 11) 11.0 M = 4 F = 7 T0 0–1.5 Negative Food Allergy 2.5 M = 2 F = 2 N/A 0–1.5 Negative (N = 4) Ulcerative Colitis (N = 3) 7.5 M = 1 F = 2 N/A 0–1.5 Negative Crohn’s Disease (N = 3) 6.6 M = 1 F = 2 T0 0–1.5 Negative * T0: Normal; T3: ﬂat destructive lesion (a: mild, b: moderate, c: total); ** N/A: non-applicable.

4.2. Determination of HLA-DQ2 and -DQ8 The determination of HLA-DQ2 and/or -DQ8 of control subjects was performed using a commercial kit (Eu-DQkit, Eurospital, Trieste, Italy). DNA was extracted from whole blood and subjected to 2 different PCR reactions. A speciﬁc genomic sequence encoding for human globin (268 bp) was used as a control, according to the manufacturers’ protocol.

4.3. Cell Culture Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation of heparinized blood using Lymphocyte Separation medium (Lonza, Walk- ersville, MD, USA). PBMCs were cultured in RPMI-1640 (Invitrogen, San Giuliano Milanese, Italy) in 6-well plates (Corning, Euroclone, Milano, Italy) at a density of 2.5 × 106/mL at ◦ 37 C in a 5% CO2 humidiﬁed atmosphere. After l h, non-adherent cells were removed. Ad- herent cells were cultured in RPMI-1640 supplemented with 10% fetal bovine serum (FBS) (GIBCO, Thermo Fisher, Milano, Italy), 1% L-glutamate (ICN), 100 U/mL streptomycin or penicillin, and 80 µL fungizone (GIBCO, Thermo Fisher, Milano, Italy). To obtain DCs, 1000 U/mL of granulocyte-monocyte colony stimulating factor (GM-CSF, RD System, Minneapolis, Minnesota, USA) and 800 U/mL of interleukin 4 (IL-4, RD System, Min- neapolis, Minnesota, USA) were added to the culture medium on Day zero and Day 3. Monocytes differentiated into immature DCs during 6 days of culture. DCs were harvested at Day 6 or 7.

4.4. Rho GTP Assay Rho activation [28] was tested by the EZ-Detect Rho Activation kit (Pierce, Rock- ford, IL, USA) following the manufacturers’ instructions. In brief, 1 × 106 DCs obtained from 5 GCD-CD, 5 GFD-CD and 5 control subjects were seeded on fibronectin for l h and then washed with ice-cold Tris-buffered solution pH 7.5. They were then lysed in Lysis/Binding/Wash Buffer (25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM MgCl2, 1% Non- idet P-40 (NP-40), 1 mM Dithiothreitol (DTT), 5% glycerol) with 1 pg/mL each of leupeptin and aprotinin, and 1 mM phenylmethylsulfonyl fluoride (PMSF) and incubated on ice for 5 min (all from Sigma-Aldrich, Milan, Italy). Cell lysates were clarified by centrifugation at 16,000× g at 4 ◦C for 15 min. Glutathione S-transferase (GST) Rhotekin Rho binding domain (RBD) (400 µg) was added to the spin cup containing the glutathione resin and then Int. J. Mol. Sci. 2021, 22, 2708 12 of 16

supernatants were incubated with GST-Rhotekin-RBD beads for 1 h at 4 ◦C. Beads were then washed 3 times with a cold lysis buffer. Bound GTP-RhoA was detected by Western blotting using monoclonal anti-RhoA antibodies (1:1000) [28,29]. All performed following manufacturer instructions (Thermo Fisher Scientiﬁc, Milan, Italy). Densitometry analysis was performed using Image J (https://imagej.nih.gov/ij) (accessed on 3 February 2021). The amount of GST-Rhotekin-RBD-bound RhoA was normalized to the total amount of RhoA in cell lysates and expressed as the GTP-RhoA/RhoA ratio. DCs from 3 CTR were treated with Nocodazole (Sigma-Aldrich, Milan, Italy) (20 µM) as a positive control [29].

4.5. PCR Total RNA was extracted from DCs from patients with CD on a GFD and CTR using TRIZOL reagent (Ambion Life Technologies, Monza, Italy). The mRNA concentration was measured using a Nanodrop spectrophotometer, and the RNA quality was analyzed using agarose gel electrophoresis in Tris/Borate/Ethylenediaminetetraacetic acid (EDTA) buffer (TBE, Sigma, Milan, Italy). The RNA (1 µg) was reverse transcribed into cDNAs using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Monza, Italy) according to the manufacturer’s protocol. Experiments were performed with approximately 40 ng of cDNA templates, according to the manufacturer’s protocol (TaqMan Gene Expression Assay), using a 7900 HT Fast Real-Time PCR system. The gene expression assay used for the ARHGAP gene was Hs00929215_m1 (Life Technologies, Monza, Italy), and the probe was located on Chromosome 3 (Chr3) exons 11–12. The expression of each gene was normalized to the expression of an endogenous housekeeping gene (Beta actin). Relative quantiﬁcation was performed using the ∆∆Ct method. SDS software (ABI, version 1.4 or 2.4) was used to analyze the raw data.

4.6. DC Spreading on Fibronectin: Actin Assay Sterile 12-mm glass coverslips (Knittel Glaser, Braunschweig Germany) were pre- coated overnight with 5 µg/mL fibronectin, then blocked for 1 h with 5% bovine serum albumin (BSA). DCs (obtained as described above) were plated onto fibronectin-coated coverslips in phosphate buffer solution (PBS) containing 5 mg/mL fetal bovine serum (FBS), 5 5 mM glucose and 0.3 MgCl2 at a concentration of 2.5 × 10 cells/mL, and incubated at 37 ◦C to evaluate their ability to spread onto a physiological substrate. After 3 h of adhering, cells were fixed in 3% paraformaldehyde (PFA, Sigma-Aldrich, Milan, Italy) and permeabilized with 0.2% Triton X-100 (Biorad, Milan, Italy). Actin rearrangements were revealed by staining with 10 µg/mL FITC-conjugated phalloidin or 0.1 mg/mL Tex-red phalloidin (Sigma-Aldrich, Milan, Italy) for 45 minutes before mounting with Mowiol (Sigma-Aldrich, Milan, Italy) (35–37). Images were acquired and processed with KS300 software (Carl Zeiss MicroImaging, Inc., Jena, Germany) [24]. Experiments were run in duplicate. Five random pictures from each duplicate were evaluated. For each subject, at least 20 adhering cells were analyzed to measure cell length.

4.7. DC Spreading on Fibronectin: Crystal Violet Assay First, 96-well plates (Nunc, Maxisorp, VWR International, Milan, Italy) were pre- coated with 50 µg/mL fibronectin in sterile water and left overnight at 4 ◦C. Wells were washed twice with sterile PBS and blocked with 5% BSA in PBS for 1 h at room temperature. All the experiments were done in duplex. After blocking, wells were washed twice before plating 10,000 DCs/well in PBS containing 5 mg/mL BSA, 5 mM glucose and ◦ ◦ 0.3 mM MgCl2, and incubated at 37 C for 1 h and 3 h. Upon incubation at 37 C, DCs were washed and stained using a solution containing 79.5% PBS, 20% methanol and 0.5% crystal violet (all from Sigma-Aldrich, Milan, Italy). The following day, wells were washed, fibronectin adhering DCs were observed, 6 random pictures were taken from each duplicate and at least 50 cells adhering cells in each picture were counted. In particular, DCs showing more than 3 small dendrites and/or 1 long dendrite were considered to have an elongated shape. For small dendrites, we defined this as a dendrite shorter than the length of Int. J. Mol. Sci. 2021, 22, 2708 13 of 16

the cell body, while long dendrites were longer than the cell’s body. At least 4 independent experiments were performed, and the average was calculated from 24 quantiﬁed ﬁelds.

4.8. Time-Lapse Experiments DCs were plated onto with wells pre-coated with ﬁbronectin at a concentration of 2.5 × 105 cells/mL and kept in a specially designed incubator (OXO-lab, Naples, Italy) monitoring the temperature and CO2 to allow live observation. More than 200 DCs in each well were ﬁlmed for 3 h; sets of frames were acquired every 18 min [41].

4.9. Flow Cytometric Analysis DCs were harvested and re-suspended in complete Roswell Park Memorial Institute (RPMI) (GIBCO, Thermo Fisher, Milan, Italy) at a concentration of 2.7–8.3 × 105 cells/mL in 24-well plates. After 1 h of incubation of DCs in absence or presence of lipopolysaccharide (LPS, Inbios, Naples, Italy) at 37 ◦C, cells were collected, washed twice and blocked for 10 min at 4 ◦C with human IgG. DCs were then incubated for 30 min at 4 ◦C with different fluorochrome-labeled antibodies (BD Bioscience, Milan, Italy), as indicated in the figure legends, diluted in Fluorescence-activated cell sorting (FACS)-PBS (BD Bio- science, Milan, Italy), washed twice using FACS-PBS and fixed with 2% paraformaldehyde (PFA) (Sigma-Aldrich, Milan, Italy). Unstained cells were used as negative control [42]. Stained cells were processed by FACS Calibur BD Pharmingen and analyzed by CellQuest (BD Biosciences, Milan, Italy).

4.10. Immunofluorescence Staining of DC Surface Markers DCs were plated onto coverslips pre-coated with fibronectin at a concentration of 2.5 × 105 cells/mL and incubated at 37 ◦C as described above. Surface staining was performed using a fluorescein isothiocyanate (FITC)-conjugated anti-HLA-DR to stain all DCs. Anti-CD14, anti-CD83, anti-CD86 and anti-CD11c phycoerythrin (PE)-conjugated antibodies (all from BD Biosciences Pharmigen, Milan, Italy) were used to perform the second staining. DCs were blocked for 10 min with human IgG at 4 ◦C, treated with the first antibody for 30 min, washed and stained with a second antibody for 30 min. Coverslips pre- coated with fibronectin were washed, fixed in 3% PFA, washed several times and mounted with Mowiol. Stained cells were observed with an Axioplan2 fluorescent microscope. Images were acquired and processed with KS300 software (Carl Zeiss MicroImaging, Inc., Jena, Germany) [24].

4.11. Statistics GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA) was used for statistical analysis and graphic representation. Statistical analyses of differences were performed using Student’s t-test or ordinary one-way ANOVA with Tukey’s multiple comparison correction, as indicated. A p-value < 0.05 was considered statistically signiﬁcant [34].

Supplementary Materials: Supplementary materials can be found at https://www.mdpi.com/1422 -0067/22/5/2708/s1. Author Contributions: V.D., M.L.G.T.E. and G.L. contributed to study conception and design; performed the spreading assays, immunoﬂuorescence and ﬂow cytometry analysis; and performed data analysis. V.D. contributed to writing the manuscript. L.S. and G.P. provided the facility to perform the video time-lapse microscopy experiments. R.T. (Rossella Tufano) performed the microscopy experiments; M.N. designed and performed the Rho activation experiments. R.T. (Riccardo Troncone) and R.A. led the clinical recruitment of patients and contributed to study conception and design. S.A. contributed to study conception and design, and supervision of the paper writing pro- cess. M.V.B. led the study conception, design and performance, and wrote the text. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Int. J. Mol. Sci. 2021, 22, 2708 14 of 16

Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the University of Naples Federico II, Protcol number 230/05 approved in 2005). Informed Consent Statement: Written informed consent was obtained from all patients or from the next of kin, caretakers, or guardians on behalf of the minor/child participants involved in our study. Acknowledgments: We would like to thank all patients that agreed to be enrolled in the study. Conflicts of Interest: The authors declare no conflict of interest. Ethical Statement: The Ethical Committee of the University “Federico II” of Naples approved the protocol of this study (ethical approval: C.E. no. 230/05). Written informed consent was obtained from all patients or from the next of kin, caretakers, or guardians on behalf of the minor/child participants involved in our study. The authors confirm that all methods were performed in accordance with the relevant guidelines and regulations.

Abbreviations

ARHGAP31 Rho GTPase-activating protein 31 BSA bovine serum albumin CD (-83, -14, -40, -86, -11c) cluster of differentiation CD celiac disease C1ORF108 C1 Orfan gene 108 DCs dendritic cells EGF epidermal growth factor EGFR epidermal growth factor receptor FBS fetal bovine serum FITC ﬂuorescein isothiocyanate GCD gluten-containing diet GFD gluten-free diet GTP guanosine triphosphate h hour/hours HLA human leukocyte antigen Ig human immunoglobulin LPP lipoma-preferred partner LPS lipopolysaccharide membrane-associated guanylate kinase, WW and PDZ MAGI2 domain-containing 2 MYO9B myosin IXB NOC nocodazole NT not treated PARD3 Par-3 family cell polarity regulator PBMCs peripheral blood mononuclear cells mo-DCs monocyte derived dendritic cells PE phycoerythrin PFA paraformaldehyde PPP2R3A protein phosphatase 2 regulatory subunit B” alpha PTPRK protein tyrosine phosphatase receptor type K RhoA Ras homolog family member A ROCK Rho-associated protein kinase Th T helper TJ tight junctions tTg tissue transglutaminase Int. J. Mol. Sci. 2021, 22, 2708 15 of 16

References 1. Jabri, B.; Sollid, L.M. Tissue mediated control of immunopathology in coeliac disease. Nat. Rev. Immunol. 2009, 9, 858–870. [CrossRef] 2. Maiuri, L.; Ciacci, C.; Ricciardelli, I.; Vacca, L.; Raia, V.; Auricchio, S.; Picard, J.; Osman, M.; Quaratino, S.; Londei, M. Association between innate response to gliadin and activation of pathogenic T cells in coeliac disease. Lancet 2003, 362, 30–37. [CrossRef] 3. Meresse, B.; Chen, Z.; Ciszewski, C.; Tretiakova, M.; Bhagat, G.; Krausz, T.N.; Raulet, D.H.; Lanier, L.L.; Groh, V.; Spies, T.; et al. Coordinated induction by IL15 of a TCR-independent NKG2D signaling pathway converts CTL into lymphokine-activated killer cells in celiac disease. Immunity 2004, 21, 357–366. [CrossRef] 4. Marsh, N.M.; Crowe, P.T. Morphology of the mucosal lesion in gluten sensitivity. Bailliere’s Clin. Gastroenterol. 1995, 9, 273–293. [CrossRef] 5. Holmgren Peterson, K.; Magnusson, K.E.; Stenhammar, L.; Fälth-Magnusson, K. Confocal laser scanning microscopy of small- intestinal mucosa in celiac disease. Scand. J. Gastroenterol. 1995, 30, 228–234. [CrossRef] 6. De Ritis, G.; Auricchio, S.; Jones, H.W.; Lew, E.J.; Bernardin, J.E.; Kasarda, D.D. In vitro (organ culture) studies of the toxicity of specific A-gliadin peptides in celiac disease. Gastroenterology 1988, 94, 41–49. [CrossRef] 7. Bjarnason, I.; Peters, T.J. In vitro determination of small intestinal permeability: Demonstration of a persistent defect in patients with coeliac disease. Gut 1984, 25, 145–150. [CrossRef] 8. Bjarnason, I.; Marsh, M.N.; Price, A.; Levi, A.J.; Peters, T.J. Intestinal permeability in patients with coeliac disease and dermati- tis herpetiformis. Gut 1985, 26, 1214–1219. [CrossRef] 9. Sander, G.R.; Cummins, A.G.; Henshall, T.; Powell, B.C. Rapid disruption of intestinal barrier function by gliadin involves altered expression of apical junctional proteins. FEBS Lett. 2005, 579, 4851–4855. [CrossRef] 10. Ciccocioppo, R.; Finamore, A.; Ara, C.; Di Sabatino, A.; Mengheri, E.; Corazza, G.R. Altered expression, localization, and phos- phorylation of epithelial junctional proteins in celiac disease. Am. J. Clin. Pathol. 2006, 125, 502–511. [CrossRef] 11. Schulzke, J.D.; Bentzel, C.J.; Schulzke, I.; Riecken, E.O.; Fromm, M. Epithelial tight junction structure in the jejunum of children with acute and treated celiac sprue. Pediatr. Res. 1998, 43, 435–441. [CrossRef] 12. Rauhavirta, T.; Lindfors, K.; Koskinen, O.; Laurila, K.; Kurppa, K.; Saavalainen, P.; Mäki, M.; Collin, P.; Kaukinen, K. Impaired epithelial integrity in the duodenal mucosa in early stages of celiac disease. Transl. Res. 2014, 164, 223–231. [CrossRef] 13. Wapenaar, M.C.; Monsuur, A.J.; van Bodegraven, A.A.; Weersma, R.K.; Bevova, M.R.; Linskens, R.K.; Howdle, P.; Holmes, G.; Mulder, C.J.; Dijkstra, G. Associations with tight junction genes PARD3 and MAGI2 in Dutch patients point to a common barrier defect for coeliac disease and ulcerative colitis. Gut 2008, 57, 463–467. [CrossRef] 14. Bailey, D.S.; Jauregi-Miguel, A.; Fernandez-Jimenez, N.; Irastorza, I.; Plaza-Izurieta, L.; Vitoria, J.C.; Bilbao, J.R. Alteration of tight junction gene expression in celiac disease. J. Pediatr. Gastroenterol. Nutr. 2014, 58, 762–767. 15. Freedman, A.R.; Price, S.C.; Chescoe, D.; Ciclitira, P.J. Early biochemical responses of the small intestine of coeliac patients to wheat gluten. Gut 1989, 30, 78–85. 16. Wilson, S.; Volkov, Y.; Feighery, C. Investigation of Enterocyte Microfilaments in Celiac Disease. In Proceedings of the 12th International Symposium on Coeliac Disease, NYC, NY, USA, 2004; McMillan, S., Feighery, C., Watson, P., Eds.; Northern Ireland: Belfast, UK, 2004; p. 100. 17. Giovannini, C.; Matarrese, P.; Scazzocchio, B.; Varí, R.; Archivi, M.D.; Straface, E.; Masella, R.; Malorni, W.; Vincenzi, D.M. Wheat gliadin induces apoptosis of intestinal cells via an autocrine mechanism involving Fas-Fas ligand pathway. FEBS Lett. 2003, 540, 117–124. [CrossRef] 18. Barone, M.V.; Gimigliano, A.; Castoria, G.; Paolella, G.; Maurano, F.; Paparo, F.; Maglio, M.; Mineo, A.; Miele, E.; Nanayakkara, M.; et al. Growth factor-like activity of gliadin, an alimentary protein: Implications for coeliac disease. Gut 2007, 56, 480–488. [CrossRef] 19. Elli, L.; Roncoroni, L.; Doneda, L.; Ciulla, M.M.; Colombo, R.; Braidotti, P.; Bonura, A.; Bardella, M.T. Imaging analysis of the gliadin direct effect on tight junctions in an in vitro three-dimensional Lovo cell line culture system. Toxicol. Vitro 2011, 25, 45–50. [CrossRef] 20. Reinke, Y.; Behrendt, M.; Schmidt, S.; Zimmer, K.-P.; Naim, H.Y. Impairment of protein trafficking by direct interaction of gliadin peptides with actin. Exp. Cell Res. 2011, 317, 2124–2135. [CrossRef] 21. Trynka, G.; Hunt, K.A.; Bockett, N.A.; Romanos, J.; Mistry, V.; Szperl, A.; Bakker, S.F.; Bardella, M.T.; Bhaw-Rosun, L.; Castillejo, G.; et al. Dense genotyping identifies and localizes multiple common and rare variant association signals in celiac disease. Nat. Genet. 2011, 43, 1193–1201. [CrossRef] 22. Almeida, R.; Ricaño-Ponce, I.; Kumar, V.; Deelen, P.; Szperl, A.; Trynka, G.; Gutierrez-Achury, J.; Kanterakis, A.; Westra, H.J.; Franke, L.; et al. Fine mapping of the celiac disease-associated LPP locus reveals a potential functional variant. Hum. Mol. Genet. 2014, 23, 2481–2489. [CrossRef][PubMed] 23. Grunewald, T.G.; Pasedag, S.M.; Butt, E. Cell Adhesion and Transcriptional Activity—Defining the Role of the Novel Protoonco- gene LPP. Transl. Oncol. 2009, 2, 107–116. [CrossRef] 24. Kumar, V.; Gutierrez-Achury, J.; Kanduri, K.; Almeida, R.; Hrdlickova, B.; Zhernakova, D.V.; Westra, H.J.; Karjalainen, J.; Ricaño-Ponce, I.; Li, Y.; et al. Systematic annotation of celiac disease loci refines pathological pathways and suggests a genetic explanation for increased interferon-gamma levels. Hum. Mol. Genet. 2015, 24, 397–409. [CrossRef] Int. J. Mol. Sci. 2021, 22, 2708 16 of 16

25. Martucciello, S.; Lavric, M.; Toth, B.; Korponay-Szabo, I.; Nadalutti, C.; Myrsky, E.; Rauhavirta, T.; Esposito, C.; Sulic, A.M.; Sblattero, D.; et al. RhoB is associated with the anti-angiogenic effects of celiac patient transglutaminase 2-targeted autoantibodies. J. Mol. Med. 2012, 90, 817–826. [CrossRef] 26. Liao, N.; Chen, M.L.; Zhao, H.; Xie, Z.F. Association between the MYO9B polymorphisms and celiac disease risk: A meta-analysis. Int. J. Clin. Exp. Med. 2015, 8, 14916–14925. 27. Nanayakkara, M.; Kosova, R.; Lania, G.; Sarno, M.; Gaito, A.; Galatola, M.; Greco, L.; Cuomo, M.; Troncone, R.; Auricchio, S.; et al. A celiac cellular phenotype, with altered LPP sub-cellular distribution, is inducible in controls by the toxic gliadin peptide P31-43. PLoS ONE 2013, 8, e79763. [CrossRef] 28. Bogunovic, M.; Ginhoux, F.; Helft, J.; Shang, L.; Hashimoto, D.; Greter, M.; Liu, K.; Jakubzick, C.; Ingersoll, M.A.; Leboeuf, M.; et al. Origin of the lamina propria dendritic cell network. Immunity 2009, 31, 513–525. [CrossRef] 29. Heasman, S.J.; Ridley, A.J. Mammalian Rho GTPases: New insights into their functions from in vivo studies. Nat. Rev. Mol. Cell Biol. 2008, 9, 690–701. [CrossRef] 30. Takesono, A.; Heasman, S.J.; Wojciak-Stothard, B.; Garg, R.; Ridley, A.J. Microtubules regulate migratory polarity through Rho/ROCK signaling in T cells. PLoS ONE 2010, 5, e8774. [CrossRef] 31. Strisciuglio, C.; Miele, E.; Wildenberg, M.E.; Giugliano, F.P.; Andreozzi, M.; Vitale, A.; Capasso, F.; Camarca, A.; Barone, M.V.; Staiano, A.; et al. T300A variant of autophagy ATG16L1 gene is associated with decreased antigen sampling and processing by dendritic cells in pediatric Crohn’s disease. Inflamm. Bowel Dis. 2013, 11, 2339–2348. [CrossRef] 32. Cardoso, S.; Cardoso-Silva, D.; Delbue, D.; Itzlinger, A.; Moerkens, R.; Withoff, S.; Branchi, F.; Schumann, M. Intestinal Barrier Function in Gluten-Related Disorders. Nutrients 2019, 11, 2325. [CrossRef] 33. Drabinska, N.; Krupa-Kozak, U.; Jarocka-Cyrta, E. Intestinal Permeability in Children with Celiac Disease after the Administration of Oligofructose-Enriched Inulin into a Gluten-Free Diet—Results of a Randomized, Placebo-Controlled, Pilot Trial. Nutrients 2020, 12, 1736. [CrossRef] 34. Linder, S.; Wiesner, C. Tools of the trade: Podosomes as multipurpose organelles of monocytic cells. Cell Mol. Life Sci. 2015, 72, 121–135. [CrossRef][PubMed] 35. Nanayakkara, M.; Lania, G.; Maglio, M.; Kosova, R.; Sarno, M.; Gaito, A.; Discepolo, V.; Troncone, R.; Auricchio, S.; Auricchio, R.; et al. Enterocyte proliferation and signaling are constitutively altered in celiac disease. PLoS ONE 2013, 8, e76006. [CrossRef][PubMed] 36. Lania, G.; Nanayakkara, M.; Maglio, M.; Auricchio, R.; Porpora, M.; Conte, M.; De Matteis, M.A.; Rizzo, R.; Luini, A.; Discepolo, V.; et al. Constitutive alterations in vesicular trafficking increase the sensitivity of cells from celiac disease patients to gliadin. Commun. Biol. 2019, 2, 190. [CrossRef][PubMed] 37. Dotsenko, V.; Oittinen, M.; Taavela, J.; Popp, A.; Peräaho, M.; Staff, S.; Sarin, J.; Leon, F.; Isola, J.; Mäki, M.; et al. Genome- Wide Transcriptomic Analysis of Intestinal Mucosa in Celiac Disease Patients on a Gluten-Free Diet and Postgluten Challenge. Cell Mol. Gastroenterol. Hepatol. 2020, S2352-345X, 30116–30118. [CrossRef] 38. Stamnaes, J.; Stray, D.; Stensland, M.; Sarna, V.K.; Nyman, T.A.; Lundin, K.E.A.; Sollid, L.M. Quantitative proteomics of coeliac gut during 14-day gluten challenge: Low-level baseline inflammation despite clinical and histological normality predicts subsequent response. medRxiv 2020.[CrossRef] 39. Freire, R.; Ingano, L.; Serena, G.; Cetinbas, M.; Anselmo, A.; Sapone, A.; Sadreyev, R.I.; Fasano, A.; Senger, S. Human gut derived-organoids provide model to study gluten response and effects of microbiota-derived molecules in celiac disease. Sci. Rep. 2019, 9, 7029. [CrossRef] 40. Dieterich, W.; Neurath, M.F.; Zopf, Y. Intestinal ex vivo organoid culture reveals altered programmed crypt stem cells in patients with celiac disease. Sci. Rep. 2020, 10, 3535. [CrossRef] 41. Thachil, E.; Hugot, J.P.; Arbeille, B.; Paris, R.; Grodet, A.; Peuchmaur, M.; Codogno, P.; Barreau, F.; Ogier-Denis, E.; Berrebi, D.; et al. Abnormal activation of autophagy-induced crinophagy in Paneth cells from patients with Crohn’s disease. Gastroenterology 2012, 142, 1097–1099.e4. [CrossRef] 42. Barone, M.V.; Troncone, R.; Auricchio, S. Gliadin peptides as triggers of the proliferative and stress/innate immune response of the celiac small intestinal mucosa. Int. J. Mol. Sci. 2014, 15, 20518–20537. [CrossRef][PubMed] 43. Sepe, L.; Ferrari, M.C.; Cantarella, C.; Fioretti, F.; Paolella, G. Ras activated ERK and PI3K pathways differentially affect directional movement of cultured fibroblasts. Cell Physiol. Biochem. 2013, 31, 123–142. [CrossRef][PubMed] 44. Barone, M.V.; Zanzi, D.; Maglio, M.; Nanayakkara, M.; Santagata, S.; Lania, G.; Miele, E.; Ribecco, M.T.S.; Maurano, F. Gliadin- mediated proliferation and innate immune activation in celiac disease are due to alterations in vesicular trafficking. PLoS ONE 2011, 6, e17039. [CrossRef][PubMed]